A Holocene Pollen Record of Persistent Droughts from Pyramid Lake, Nevada, USA

Home , Hidden Cave

University of Nebraska - Lincoln DigitalCommons@University of Nebraska - Lincoln

USGS Staff -- Published Research US Geological Survey

2004

A Holocene pollen record of persistent droughts from Pyramid Lake, Nevada, USA

Scott A. Mensing University of Nevada - Reno, [email protected]

Larry Benson University of Colorado at Boulder, [email protected]

Michaele Kashgarian Lawrence Livermore National Laboratory

Steve Lund University of Southern California, [email protected]

Follow this and additional works at: https://digitalcommons.unl.edu/usgsstaffpub

Part of the Geology Commons, Oceanography and Atmospheric Sciences and Meteorology Commons, Other Earth Sciences Commons, and the Other Environmental Sciences Commons

Mensing, Scott A.; Benson, Larry; Kashgarian, Michaele; and Lund, Steve, "A Holocene pollen record of persistent droughts from Pyramid Lake, Nevada, USA" (2004). USGS Staff -- Published Research. 783. https://digitalcommons.unl.edu/usgsstaffpub/783

This Article is brought to you for free and open access by the US Geological Survey at DigitalCommons@University of Nebraska - Lincoln. It has been accepted for inclusion in USGS Staff -- Published Research by an authorized administrator of DigitalCommons@University of Nebraska - Lincoln. Quaternary Research 62 (2004) 29–38 www.elsevier.com/locate/yqres

A Holocene pollen record of persistent droughts from Pyramid Lake, Nevada, USA

Scott A. Mensing,a,* Larry V. Benson,b Michaele Kashgarian,c and Steve Lundd

a Department of Geography, University of Nevada, Reno, NV 89557, USA b U.S. Geological Survey, 3215 Marine Street, Boulder, CO, 80303, USA c Lawrence Livermore National Laboratory, PO Box 808, Livermore, CA, 94550, USA d Department of Earth Sciences, University of Southern California, Los Angeles, CA, 90089, USA Received 19 August 2003

Abstract

Pollen and algae microfossils preserved in sediments from Pyramid Lake, Nevada, provide evidence for periods of persistent drought during the Holocene age. We analyzed one hundred nineteen 1-cm-thick samples for pollen and algae from a set of cores that span the past 7630 years. The early middle Holocene, 7600 to 6300 cal yr B.P., was found to be the driest period, although it included one short but intense wet phase. We suggest that Lake Tahoe was below its rim for most of this period, greatly reducing the volume and depth of Pyramid Lake. Middle Holocene aridity eased between 5000 and 3500 cal yr B.P. and climate became variable with distinct wet and dry phases. Lake Tahoe probably spilled intermittently during this time. No core was recovered that represented the period between 3500 and 2600 cal yr B.P. The past 2500 years appear to have had recurrent persistent droughts. The timing and magnitude of droughts identified in the pollen record compares favorably with previously published y18O data from Pyramid Lake. The timing of these droughts also agrees with the ages of submerged rooted stumps in the Eastern Sierra Nevada and woodrat midden data from central Nevada. Prolonged drought episodes appear to correspond with the timing of ice drift minima (solar maxima) identified from North Atlantic marine sediments, suggesting that changes in solar irradiance may be a possible mechanism influencing century-scale drought in the western Great Basin. D 2004 University of Washington. All rights reserved.

Keywords: Pyramid Lake; Nevada; Great Basin; Pollen; Drought; Holocene; Climate change

Introduction cores from lakes with high depositional rates. Benson et al. (2002) published a 7600-year record of y18O of endogenic In western North America, periods of extended drought carbonates and magnetic susceptibility from Pyramid Lake, have significant effects on human populations. Evidence for Nevada sediment cores that documented multidecadal and recurrent droughts that may have lasted a century or longer multicentennial droughts in the northern Sierra Nevada. In comes from submerged stumps found in lakes along the this paper, we present a high-resolution pollen record recon- eastern Sierra Nevada Range (Harding, 1935, 1965; Lind- structed from the same cores that documents vegetation stro¨m, 1990; Stine, 1990). On shorter time scales, river flow change associated with persistent drought in the Holocene. reconstructed from tree-ring widths indicates periods of as Pyramid Lake (1160 m surface elevation) is a closed-basin many as 50 consecutive years when annual runoff of the lake located in the rainshadow of the Sierra Nevada (Fig. 1). Sacramento River remained below the 94-yr historical mean The primary tributary is the Truckee River, which has its (Meko et al., 2001). In order to characterize the potential for headwaters on the eastern slope of the Sierra Nevada. One- future long-term droughts, high-resolution analyses of con- third of the Truckee’s flow originates from Lake Tahoe. In the tinuous proxy-climate records for the past few thousand headwaters of the Truckee River, annual precipitation ranges years are needed. One source for this evidence is sediment from 1000 mm at Donner Lake to 805 mm at Lake Tahoe (National Climatic Data Center, 2003). At Pyramid Lake, * Corresponding author. Fax: (775) 784-1058. annual precipitation at Sutcliffe, on the western shore, aver- E-mail address: [email protected] (S.A. Mensing). ages 200 mm, whereas on the southeastern shore it averages

Fig. 1. Location maps of the study area and sites referred to in this paper. Abbreviations are: BL, Bunker Lake; BM, Balsam Meadow; DP, Diamond Pond; EM, Exchequer Meadow; GSL, Great Salt Lake; HC, Hidden Cave; KC, Krammer Cave; LT, Lake Tahoe; LV, Little Valley; ML, Mono Lake; OL, Owens Lake; PL, Pyramid Lake; RM, Ruby Marsh; SL, Swamp Lake; TR, Toiyabe Range, and WL, Walker Lake.

160 mm. Most precipitation in the Sierra Nevada falls during monticola (western white pine), P. albicaulis (whitebark the winter months (Houghton et al., 1975). pine), Abies magnifica (red fir), and Tsuga mertensiana Pyramid Lake lies along the ecotone between the shad- (mountain hemlock) dominate the landscape (Lanner, scale and sagebrush vegetation zones (Billings, 1949). 1984). Understory shrubs include Quercus vaccinifolia Shadscale typically dominates where annual precipitation (huckleberry oak), Ceanothus velutinus (tobacco brush), is <150 mm, whereas sagebrush dominates where precipi- and Arctostaphylos spp. (manzanita). Montane riparian spe- tation is >180 mm. Vegetation on the western, wetter side of cies include Alnus tenuifolia (mountain alder) and Betula Pyramid Lake is dominated by Artemisia tridentata (big occidentalis (water birch), in addition to Salix and Populus sagebrush) and Ephedra nevadensis (Mormon tea). Also spp. Cercocarpus ledifolius (curlleaf mountain mahogany) present are Purshia tridentata (bitterbrush), Tetradymia spp. grows on dry rocky slopes above the sagebrush zone between (horsebrush), and Eriogonum spp. (buckwheat). Vegetation 1500 and 3000 m elevation. on the drier eastern shore is dominated by members of the Chenopodiaceae family, including Atriplex confertifolia (shadscale), Atriplex canescens (four-wing saltbush), Krash- Methods cheninnikovia lanata (winterfat), and Grayia spinosa (spiny hopsage), and the Sarcobataceae family, including Sarcoba- Three cores were recovered from the deepest part of tus vermiculatus (greasewood),. The area surrounding Pyr- Pyramid Lake (107 m) in 1997 and 1998: a 0.59-m-long amid Lake is a treeless landscape, but scattered Juniperus box core (PLB98-2) that recovered undisturbed surface osteosperma (Utah juniper) are present at higher elevations sediments, a 5.35-m-long piston core (PLC97-1) and a and an isolated pocket of Pinus jeffreyi (Jeffrey pine) can be 6.34-m-long piston core (PLC98-4). Age control is de- found in the Virginia Mountains west of the lake. Salix spp. scribed in Benson et al. (2002). Accelerator mass spec- (willows) and Populus fremontii (cottonwood) are common trometry (AMS) radiocarbon dates were obtained on pine along the Truckee River. pollen (Mensing and Southon, 1999). The pollen dates The nearest mixed conifer forest is approximately 50 km were about 650 years older than the age model derived west in the Sierra Nevada, above f1650 m elevation, and is from paleomagnetic secular variation (PSV). Benson et al. characterized by Pinus jeffreyi, P. ponderosa (ponderosa (2002) found that Holocene-age sediments are missing pine), Abies concolor (white fir), and Calocedrus decurrens from the margins of Pyramid Lake. Surface sediments from (incense cedar). In the Lake Tahoe watershed, at elevations shallow water (ca. 7 m depth) date to f12,000 cal yr B.P. >2300 m, trees including Pinus contorta (lodgepole pine), P. The old radiocarbon dates on pollen indicate some rework- S.A. Mensing et al. / Quaternary Research 62 (2004) 29–38 31 ing of Pleistocene-age pollen from the lake margins into the We tested the validity of the A/C ratio on human- deep basin. We assume that f10% of the fine-sediment induced lake-level decline in Pyramid Lake. Hydrologic fraction in the deep basin has been reworked from the lake modeling shows that during the past century, the surface margins (Benson et al., 2002). elevation of Pyramid Lake would have dropped by only A total of one hundred nineteen 1-cm-thick samples 3.5 m due to climate effects alone (Benson et al., 2002).In were taken for pollen analysis; 14 from core PLB98-2, 56 fact, the surface elevation of Pyramid Lake has dropped from core PLC97-1, and 49 from core PLC98-4. Pollen >20 m as a result of water diversions from the Truckee preparation followed standard methods (Faegri and Iversen, River initiated in 1906. We analyzed pollen samples from 1985). A known quantity of Lycopodium spores was added the historic period (200–0 cal yr B.P.) at an average to each sample to calculate pollen concentration (Stock- interval of 13 yr to test whether the A/C ratio responded marr, 1971). Pollen was identified to the lowest possible as predicted to this human-induced lake-level decline or taxonomic level using modern reference material and apparent drought. published pollen keys (Kapp et al., 2000; Moore and Webb, 1978). TCT (Taxodiaceae/Cupressaceae/Taxaceae) pollen was identified, but not used in the interpretation because Results it can represent two ecologically distinct taxa, Calocedrus decurrens, a montane species, or Juniperus osteosprema,a In this paper, results are discussed for several periods in desert species. Algae also were identified, including the Holocene. The historic period (200–0 cal yr B.P.) is Botryococcus and four species of Pediastrum (Jankovska´ discussed first, and then prehistoric periods in the Holocene and Koma´rek, 1982). (7600–200 cal yr B.P.). A minimum of 400 pollen grains were counted in each sample. Pollen percentages were calculated from the sum The historic period (200–0 cal yr B.P.) of terrestrial pollen and spores. Pollen accumulation rates were calculated by dividing the pollen concentration by the Core PLB98-2 spans the period from 1860 to 1998 A.D., number of years represented by the sample. Algae were encompassing nearly the entire historical record. The pollen counted concurrently with pollen, and percentages were diagram can be divided into two zones, prediversion and calculated from the sum of total algae. Pollen zonation was postdiversion (Fig. 2). Pollen accumulation rates support the interpreted from a constrained single-link dendrogram percentage data throughout the record. Pinus pollen aver- created using a cluster analysis program modified from aged 50% between 1820 and 1865 A.D. and then decreased Birks and Gordon (1985), using the 14 most common taxa to 35% by 1895 A.D. Pinus remained relatively low (>93% of the total pollen sum for each sample). Pollen (f35%) throughout the past 100 yr. The initial decrease spectra for each level were compared with every other could be due to either an actual decrease in pine pollen level in the record using the squared-chord distance dis- production in the Sierra Nevada at the end of the Little Ice similarity coefficient with a critical value of 0.12 (Over- Age, or removal of trees through clear-cutting in the upper peck et al., 1985) to test whether pollen assemblages watershed. Lumbering began in the mid-1850s A.D., within zones were distinct from pollen assemblages in boomed between 1860 and 1890 A.D., and slowed by the different zones. turn of the century (Wilson, 1992). Diversions appear to We used the ratio between Artemisia and Chenopodia- have affected riparian taxa such as Salix and Alnus, as these ceae pollen as a proxy for lake-level change. Billings (1949) two pollen types decrease with the initiation of diversion noted that species of the shadscale association (such as and then increase in association with increased runoff Atriplex confertifolia, Atriplex canescens, and Sarcobatus between 1970 and 1990 A.D. vermiculatus) typically occupy both moderately saline playa Chenopodiaceae pollen began to increase in relation to soils and xeric habitats. We hypothesize that during periods Artemisia pollen about 1950 A.D., approximately 40 yr of drier climate when Pyramid Lake shrank exposing playa after diversion began. The lag in pollen production is surfaces, individuals of the shadscale association would consistent with our expectation of the time needed for likely have dispersed into this habitat and increased in plants to colonize the newly exposed surface and become number on the landscape. Conversely, periods of wetter mature. Percent Artemisia pollen decreased steadily after climate would have favored expansion of sagebrush and water diversion began. This signal is likely a response to rising lake levels would have eliminated playa habitat. Thus, the lake edge withdrawing further from the sagebrush the ratio between percent Artemisia and Chenopodiaceae zone. Although Artemisia is wind-pollinated, the pollen pollen (A/C ratio) should provide a proxy for lake-level does not disperse far, typically traveling <2 km from its change. We expect this signal to be lagged due to the time point of origin (Solomon and Silkworth, 1986). The A/C required for plant succession onto the playa. Previous ratio (Artemisia/Chenopodiaceae) decreased as predicted studies by Mensing (2001), Wigand (1987), and Byrne et for lowered lake levels (Fig. 2), confirming that this ratio al. (1979) have all interpreted the A/C ratio as a relative can be used as a proxy for lowered lake levels due to measure of available moisture in the Great Basin. reduced river flows caused by extended drought in the 32 S.A. Mensing et al. / Quaternary Research 62 (2004) 29–38

Fig. 2. Pollen percentage diagram for core PLB98-2 (box core). Lake level curve graphs the average annual surface elevation of Pyramid Lake. The ratio of Artemisia/Chenopodiaceae + Sarcobatus (A/C ratio) is calculated as (a-c)/(a+c), where ‘‘a’’ represents percent Artemisia and ‘‘c’’ represents percent Chenopodiaceae + Sarcobatus. Positive values represent increased Artemisia (wetter climate) and negative values represent increased Chenopodiaceae (drier climate). Lines to the right of filled curves are 5Â exaggerations. absence of water diversions. The A/C ratio has approxi- tions or the invasion of extralocal species. Pollen and algae mately a 40- to 50-yr lag compared to the lake-level stratigraphy for the four pollen zones is described in the curve during the historic period. Since the A/C ratio only following section. records droughts of sufficient duration and intensity that lake level drops for periods of >50 yr, we infer that minima in this ratio represent the most severe drought Interpretation and comparison with other records episodes. 7530 to 6300 cal yr B.P. (zone 1A) The Holocene before the historic period (7530–200 cal yr B.P.) The Pyramid Lake pollen data record high percentages of Chenopodiaceae in relation to Artemisia, high percen- Core PLC97-1 overlaps the box core, PLB98-2, and tages of Ambrosia, and low percentages of Pinus pollen in spans the period from 2640 to f130 cal yr B.P. Core zone 1A between 7530 and 6300 cal yr B.P. The A/C ratio PLC98-4 spans the period from 7600 to 3430 cal yr B.P. (Fig. 3) suggests expansion of drought-tolerant species, There is a gap in the record between 3430 and 2740 cal yr such as Atriplex confertifolia, A. canescens, and Sarcoba- B.P. (Fig. 3) (Benson et al., 2002). The sediment accumu- tus vermiculatus. The historic pollen record (Fig. 2) lation rate for PLC98-4 averaged 0.15 cm-1yr-1 and that for indicates that these species increase when lake-level drops PLC97-1 averaged 0.21 cm-1yr-1. The interval between and new shoreline is exposed. Ambrosia, a colonizer of pollen samples in PLC98-4 averages 85 yr (range 43 to open ground, probably also expanded in response to lower 155 yr) and that in PLC97-1 averages 44 yr (range 11 to 80 lake levels. Algae data also support the interpretation of a yr). Pollen analysis began in samples at 7530 cal yr B.P. shallower lake. Pediastrum, predominates when lake level Although the cluster analysis identified distinct group- drops, as can be seen in the historic period, and is the ings (Fig. 3), the squared-chord distance dissimilarity dominant algae in zone 1A (Fig. 4). analysis showed that few of the pollen spectra within each Magnetic susceptibility data from core PLC98-4 are zone were significantly different from the pollen spectra of highest in zone 1A and support the interpretation of lower other zones. This indicates that the magnitude of Holocene lake levels between 7600 and 6300 cal yr B.P. (Fig. 4). climate change was not sufficient to lead to local extinc- Benson et al. (2002) stated that magnetic susceptibility S.A. Mensing et al. / Quaternary Research 62 (2004) 29–38 33

Fig. 3. Pollen percentage diagram of the most common pollen types for cores PLB98-2, PLC97-1, and PLC98-4. The ratio of Artemisia/Chenopodiaceae + Sarcobatus (A/C ratio) is calculated as (a-c)/(a+c), where ‘‘a’’ represents percent Artemisia and ‘‘c’’ represents percent Chenopodiaceae + Sarcobatus. Positive values represent increased Artemisia (wetter climate) and negative values represent increased Chenopodiaceae (drier climate). Dark fill in the A/C ratio represents periods interpreted to be extensive droughts. y18O is a 40-yr running average and wet–dry oscillations are numbered 1–47 following Benson et al. (2002). Minimum values for y18O indicate the termination of a wet period, beginning of a dry period. Dashed lines indicate the suggested correlation between inferred droughts in the y18O and pollen records. Submerged stump radiocarbon ages are represented by geometric symbols: solid circles are from various locations in Lake Tahoe (Lindstro¨m, 1990); open circle is Rubicon Point, Lake Tahoe (Benson et al., 2002); square is Fallen Leaf Lake (Benson et al., 2002); and triangles are from Mono Lake (Stine, 1990; 1994). Lines to the right of filled curves are 5Â exaggerations. increased in the lake’s depocenter as the lake shrank, i.e., Lake Tahoe with a death date of 6300 cal yr B.P. more magnetite eroded from the basin margin reached the (Lindstro¨m, 1990). This date correlates with a strongly deepest part of the lake. negative A/C ratio and the ending of a significant drought The y18O record from Pyramid Lake supports the indicated by the y18O data (oscillation 7, Fig. 3). existence of periodic extended droughts. Benson et al. Studies from lakes on the western Sierra Nevada Range (2002) identified seven wet–dry oscillations between have interpreted the pollen record to suggest that the early 7600 and 6300 cal yr B.P. (Fig 3).Minimainy18O Holocene was warm and dry, and that cooler moister con- indicate the termination of wet periods and the beginning ditions began earlier than is seen in the Pyramid Lake record. of droughts. Oscillations 3–7 indicate extended periods of Pollen evidence of a shift from Artemisia to Pinus dominance declining lake levels, which correspond to negative values at Balsam Meadow has been interpreted as a shift from a dry in the A/C ratio. The pollen-sampling interval is not climate to wetter climate by f7800 cal yr B.P. Davis et al. sufficient to identify a one-to-one correspondence between (1985). Similar evidence from Exchequer Meadow, south of the two proxy records throughout this time period. At Balsam Meadow, has also been interpreted as a shift from dry Mono Lake, a change to lower sedimentation and pollen to wetter climate by f7800 cal yr B.P. (Davis and Moratto, accumulation rates at 7500 cal yr B.P. has been interpreted 1988). At Swamp Lake in Yosemite National Park, north of as decreased stream discharge into the lake (Davis, 1999), Balsam Meadow (Fig. 1) a change from high percentages of an interpretation consistent with the Pyramid Lake record. Quercus pollen to increased Abies pollen at about 7350 cal yr Submerged stumps in Lake Tahoe have been interpreted B.P. was interpreted as a shift at this time from warm dry as evidence for century-long droughts (Harding, 1965; climate to cooler and/or moister climate (Smith and Ander- Lindstro¨m, 1990). Ring counts on submerged stumps son, 1992). Still further north, evidence from Bunker Lake indicate minimum ages of 100 yr. The deepest and oldest indicates a warm and dry climate persisting until f6800 cal stump found is rooted 3.68 m below the natural sill of yr B.P. (Edlund, 1996). A possible explanation for these 34 S.A. Mensing et al. / Quaternary Research 62 (2004) 29–38

Fig. 4. Algae percent diagram and magnetic susceptibility. Magnetic susceptibility is represented as a 40-yr running average (Benson et al., 2002). results is that the timing and duration of early and middle 6300 to 5000 cal yr B.P. (zone 1B) Holocene drought varied spatially, with warm dry climate persisting longer in the northern Sierra Nevada and the Pollen and algae indicate a shift towards wetter climate eastern side of the range. and increased lake level beginning f6300 cal yr B.P. The Although the period from 7530 to 6300 cal yr B.P. appears A/C ratio increased indicating an increase in sagebrush to have been the driest in our record for the western Great relative to saltbush species. Mean percent Ambrosia pollen Basin, it included a period of wetter climate. The strongest decreased from 2% in zone 1A to 1% in zone 1B and Pinus evidence for this variability comes from the y18O record, increased from 35% to 40% (Fig. 3). Percent Pediastrum which shows a substantial excursion towards lower values remained high relative to Botryococcus, but the mean between 7000 to 6400 cal yr B.P. (Fig. 3). High temperatures declined from 79% Pediastrum in zone 1A to 64 % in can produce low y18O values due to decreased isotopic zone 1B (Fig. 4). Between 5200 and 5000 cal yr B.P. there fractionation between carbonate and water, but the values is a sharp decline in the A/C ratio coinciding with a sharp seen here would have required a temperature increase of increase in Pediastrum from 25% to 82%. This change 19jC. An increase of this magnitude would have completely correlates with an inferred drought in the y18Odata changed the local plant species, and we see no evidence of (oscillation 17, Fig. 3). this. An alternative explanation is that a pulse of fresh water Western Sierra Nevada pollen records all agree that entered the shallow, hydrologically closed lake. Benson et al. climate had become wetter and/or cooler by 6300 cal yr (2002) suggest that a short but possibly intense wet phase B.P. (Davis et al., 1985; Davis and Moratto, 1988; Smith raised the lake level between 7000 and 6800 cal yr B.P. This and Anderson, 1992; Edlund, 1996). short but significant wet phase also is recorded by a higher A set of Lake Tahoe stumps are rooted between 0 and 1.3 lake-level in the Great Salt Lake (Madsen et al., 2001). m below the natural sill with death dates ranging from 5600 Between 6750 and 6700 cal yr B.P., percent Botryococcus to 5000 cal yr B.P. (Lindstro¨m, 1990). The pollen evidence doubled from 20% to 41%, possibly in response to an influx indicates a drought occurred between 5200 and 5000 cal yr of fresh water (Fig. 4). The A/C ratio varies but does not B.P. (Fig. 3). It is unclear why the pollen data do not indicate a shift to wetter climate. One possible explanation for indicate an extended dry period between 5600 and 5200 this is that if the lake had a much smaller surface area, the cal yr B.P., but both the pollen and y18O data suggest that freshwater input could have a significant effect on y18O, yet the most severe drought occurred between 5200 and 5000 the lake may have still been well below modern levels with a cal yr B.P. This drought is not identified in western Sierra large expanse of saltbush species around the shoreline. Nevada pollen records (Davis et al., 1985; Davis and S.A. Mensing et al. / Quaternary Research 62 (2004) 29–38 35

Moratto, 1988; Smith and Anderson, 1992) or the Mono logical evidence from Hidden Cave (Wigand and Meh- Lake pollen record (Davis, 1999) but the sampling resolu- ringer, 1985) and Kramer Cave (Hattori, 1982) indicate tion of these studies (ca. 250–500 yr between samples) reoccupation of these sites by 4200 cal yr B.P., and precludes identification of events at the centennial scale. collection of wetland plants associated with expansion of the nearby marshes. At Diamond Pond, abundant aquatic 5000 to 3400 cal yr B.P. (zone 2) plants indicate deep water and an extensive marsh system between 4100 and 3900 cal yr B.P. (Wigand, 1987). Beginning at 5000 cal yr B.P. there is an increase in The Pyramid Lake pollen and y18O records indicate a montane tree pollen (Fig. 3). Pinus increases to its highest short dry phase between 3900 and 3800 cal yr B.P. followed average (48%) in zone 2, but is variable, ranging from 30 to by a wet phase between 3800 and 3400 cal yr B.P. Stine 65%. Average percent Abies increases from 1 to 2%. Both (1990) found evidence suggesting a highstand of Mono Chenopodiaceae and Artemisia decline slightly, but like Lake at 3770 cal yr B.P. The pollen record from Mono Lake Pinus, are highly variable throughout the zone. This vari- has the highest percentage of Pinus at this time, supporting ability is reflected in the A/C ratio. Botryococcus becomes the interpretation of a wetter climate (Davis, 1999). Between the dominant algae in zone 2 (Fig. 4). f4000 and 3300 cal yr B.P., the upper treeline on Sheep There is evidence for a wetter climate across the northern Mountain in the White Mountains lowered 100 m, probably Great Basin beginning about 5000 cal yr B.P. We suggest associated with cooler summer temperatures (LaMarche, that it was highly variable, with distinct wet and dry phases. 1973). Madsen et al. (2001) note that the Great Salt Lake The Pyramid Lake pollen record shows an initial wet phase expanded to reach the Nevada/Utah border and refer to this from 5000 to 4700 cal yr B.P (Fig. 3). The increase in Pinus major wet event as the most dramatic climate change in the and Abies pollen is probably associated with increased Holocene. Pollen and macrofossil evidence from Diamond fluvial transport. Abies is a large, poorly dispersed pollen Pond indicate expansion of juniper woodland into sagebrush type and probably only reaches Pyramid Lake via the and shadscale zones, supporting the interpretation of cooler, Truckee River. The date 5000 cal yr B.P. marks the final wetter conditions dominated by increased winter precipita- death date of rooted stumps in Lake Tahoe, suggesting that tion (Wigand, 1987). In southern Oregon, cool and wet lake level rose, drowned trees, and increased flow in the conditions persisted until f2000 cal yr B.P., whereas, in Truckee River. There also is a shift in algae from Pedias- central Nevada, they ended by 2500 cal yr B.P. Pollen trum to Botryococcus, indicating that Pyramid Lake deep- records in the western Sierra Nevada generally indicate ened. Taken as a whole, the data indicate that middle intensified cooling after about 3900 cal yr B.P. (Smith and Holocene aridity ended about 5000 cal yr B.P. Anderson, 1992). Grayson (1993) notes that after 5000 cal yr B.P. archeo- Benson et al. (2002) found that Pyramid Lake y18O logical sites increase in abundance in the Great Basin, and increased from 3430 cal yr B.P. to 2740 cal yr B.P. (Fig. 3). at Diamond Pond, Oregon, Wigand (1987) found indica- They interpreted the increase as evidence that Pyramid Lake tions of a high water table beginning f5000 cal yr B.P. At had remained isotopically closed between 5000 and 3430 cal Little Valley, 20 km east of Lake Tahoe, Abies and Alnus yr B.P. and that continued dry Sierran climate prevented Lake pollen increase, indicating a wetter climate (Wigand and Tahoe from spilling throughout most of this period. The Rhode, 2002). In central Nevada, the pollen record from pollen evidence suggests that Lake Tahoe probably did spill Ruby Marsh indicates a decline in shadscale (Chenopodia- intermittently between 5000 and 3430 cal yr B.P., but the ceae) and deeper water after 5000 cal yr B.P. (Thompson, climate was not wet enough for Lake Tahoe to spill contin- 1992). In Utah, the return of cooler and wetter conditions uously during this period. After 3430 cal yr B.P., increased resulted in freshening of the Great Salt Lake, allowing the precipitation probably caused Lake Tahoe to fill, raising the survival of the moderately saline-tolerant Utah chub and level of Pyramid Lake. Since about 3430 cal yr B.P., Lake increased diversity in waterfowl (Madsen et al., 2001).In Tahoe has probably overflowed most years, maintaining a the White Mountains, treeline advanced upslope beginning deeper Pyramid Lake than that which existed during the about 5000 cal yr B.P. (LaMarche, 1973). That study found middle Holocene. Tausch et al. (2004) reviewed pollen, an unusually large number of trees dating to the period woodrat midden, treeline, and lake-level evidence for the between 4750 and 4550 cal yr B.P. and interpreted this to Great Basin and identified the period from 3500 to 2600 cal yr indicate unusual summer warmth and relatively high B.P. as much cooler and wetter than any time during the amounts of precipitation. previous 4000 yr. Unfortunately, most of this period is absent This initial wet phase beginning at f5000 cal yr B.P. from the core record described in this study. was followed by a period of drier climate at Pyramid Lake from 4700 to 4300 cal yr B.P. indicated by lower values in 2750 to 200 cal yr B.P. (zone 3) the A/C ratio and higher values in the y18O record (oscillations 20–22). This is followed by a wetter period from Artemisia averages 18% and Pinus averages 40% in zone 4300 to 3900 cal yr B.P. This wet period is identified in 3 (Fig. 3). Chenopodiaceae pollen averages 12% throughout several Great Basin records. Paleoecological and archeo- zone 3, but has peaks between 16 and 21% between 2500 36 S.A. Mensing et al. / Quaternary Research 62 (2004) 29–38 and 2000 cal yr B.P. (Fig. 3). The A/C ratio includes a series here represents dates on outer rings (death dates) from rooted of minima associated with peak values of Chenopodiaceae. stumps and trees. The drought between 1500 and 1250 cal yr We infer these minima to represent regional droughts when B.P. corresponds with a date of 1240 F 40 14CyrB.P. Pyramid Lake was low. (f1200 cal yr B.P.) on a stump recovered from Rubicon The last 2500 yrs are notable for the recurrence of Point, Lake Tahoe (Benson et al., 2002). The drought persistent droughts. Stine (1994) suggested that submerged between 800 and 725 cal yr B.P. corresponds with a date of stumps at Mono Lake provided evidence for century long 850 F 40 14C yr B.P. (f750 cal yr B.P.) on several sub- droughts in the western Great Basin. The Pyramid Lake merged trees in Fallen Leaf Lake (Benson et al., 2002), and a pollen record supports this conclusion and provides evidence series of stumps from Mono Lake that date the termination of for a series of extended droughts in the western Great Basin the Simis Ranch low stand at f700 cal yr B.P. (Stine, 1990). (LaMarche, 1973; Meko et al., 2001; Miller et al., 2001). The drought between 600 and 450 cal yr B.P. corresponds The y18O record identifies 18 wet/dry oscillations (30–47) with a second series of Mono Lake stumps which indicate the over the past 2750 yr (Fig. 3). Multiple processes influence termination of the 10-mile Road low stand f550 cal yr B.P. the amplitude of y18O oscillations, such that the absolute (Stine, 1990). intensity of wet and dry periods is difficult to interpret; A stream flow reconstruction for the Sacramento River however the timing and duration of events can be accurately (Fig. 1) developed from tree-rings, indicated that in the past characterized (Benson et al., 2002). Twelve minima in the A/ 1100 yr, the period between 1350 and 1400 A.D. (600 to 550 C ratio correlate with dry oscillations in the y18O lagged 40– cal yr B.P.) was the driest 50-yr period and between 1140 and 65 yr. A 65-yr lag is 15–25 years longer than the lag 1160 A.D. (f800–780 cal yr B.P.) was the driest 20-yr identified from the historic period in the box core, but the period (Meko et al., 2001). At Walker Lake, a large closed- sampling interval in core PLC97-1 averages 44 yr, precluding basin lake in the western Great Basin comparable to Pyramid our ability to identify a shorter-term lag. In most cases, the Lake, the y18O data indicate generally drier conditions during beginning of droughts in the y18O record (shift from lower to the period of 950 to 590 cal yr B.P. with century-scale drought higher y18O values) corresponds to lower A/C values (shift to events terminating at 750 and 590 cal yr B.P. (Yuan et al., more Chenopodiaceae pollen) with an average lag in the 2004). Thus, the timing of major droughts identified in the pollen record of 50 yr. Between 2500 and 2000 cal yr B.P., the A/C ratio repeatedly achieves minima comparable to those recorded during the middle Holocene and equal to those in the historic period when Pyramid Lake has fallen 20 m below the historic high. These minima correspond with oscillations 30 to 35 in the y18O record (Fig. 3) and support other data that indicate an extended dry period between 2500 and 2000 cal yr B.P. (Tausch et al., 2004). Woodrat middens near streams in the Toiyabe Range in central Nevada have low diversity between 2500 and 2000 cal yr B.P., indicating local extinction of riparian species during a severe drought. Geomorphic evidence from the same region interprets active deposition on alluvial fans associated with loss of hillslope vegetation (Miller et al., 2001). Wet sedge meadows in the Toiyabe were converted to dry grassy flats about 2100 cal yr B.P. (Tausch et al., 2004), and between 2500 and 2400 cal yr B.P. treeline lowered by f30 m on Campito Mountain in the White Mountains. LaMarche (1973) speculated that this change was in response to drier conditions. A sediment core from Mono Lake contains sand layers and increased Arte- misia pollen at 2400 cal yr B.P, indicating lowered lake level and a dry climate (Davis, 1999). Sedimentary evidence of subaerially desiccated clays and stream gravels have been interpreted as a Mono Lake low stand with a provisional date of 1800 cal yr B.P. (Stine, 1990). Three more droughts are inferred from the A/C ratio, from 1500 to 1250, 800 to 725, and 600 to 450 cal yr B.P. (Fig. 3). Fig. 5. Pyramid Lake A/C ratio and North Atlantic drift ice (Bond et al., The end date of each drought corresponds with the dates of 2001) records. The A/C ratio is represented as a 200-yr running average. submerged rooted stumps recovered from lakes in the eastern Gray bands indicate periods of possible correlation between inferred Sierra Nevada and western Great Basin. All evidence cited droughts in the pollen record and periods of reduced ice rafted debris. S.A. Mensing et al. / Quaternary Research 62 (2004) 29–38 37

Pyramid Lake pollen data agrees with other regional records structed stream-flow records. Century-scale droughts that of drought reconstructed from lake-level data, tree-ring data, occurred during the past 5000 yr may be related to changes and y18O data. in solar irradiance.

Comparison with the North Atlantic ice drift record Acknowledgments Bond et al. (2001) reconstructed a record of drift ice in the North Atlantic from petrologic tracers found in marine We thank Robin Tausch, Kelly Redmond, and Peter sediment cores. They correlated tracers with the production Wigand for helpful discussion on Great Basin climate rates of two cosmogenic nuclides related to solar activity, history and reviews of an earlier version of this manuscript. 14 10 Cand Be, and found that over the past 12,000 yr, The final version of this manuscript was improved by decreases in drift ice abundance corresponded to increased comments from Emi Ito, Marith Reheis, and two solar output. General circulation models have shown that anonymous reviewers. Alan Hayvaert and Bob Richards small changes in solar forcing, such as those during the collected the box core. Gary Johnson provided spatial data Maunder Minimum, have resulted in large regional temper- for Fig. 1. The Pyramid Lake Paiute Tribe generously ature changes across the Northern Hemisphere continents provided access to Pyramid Lake. The senior author’s work (Shindell et al., 2001). We compared the pollen record of was supported by a sabbatical leave grant from the droughts from Pyramid Lake with the stacked petrologic University of Nevada, Reno. record of North Atlantic drift ice (Fig. 5). In the past 7600 yr, nearly every occurrence of a shift from ice maxima (reduced solar output) to ice minima (increased solar output) References corresponded with a period of prolonged drought in the Pyramid Lake record. Benson, L., Kashgarian, M., Rye, R., Lund, S., Paillet, F., Smoot, J., Kester, Reduced solar irradiance has been correlated with C., Mensing, S., Meko, D., Lindstrom, S., 2002. Holocene multidecadal droughts in the Yucatan Peninsula (Hoddel et al., 2001) and multicentennial droughts affecting Northern California and Nevada. and lake level changes in equatorial Africa, and drift-ice Quaternary Science Reviews 21, 659–682. cycles correspond with temperature cycles in the Sargasso Billings, W., 1949. The shadscale vegetation zone of Nevada and Eastern Sea (Bond et al., 2001). Solar forcing has been suggested as California in relation to climate and soils. The American Midland Na- turalist 42, 87–109. a possible mechanism for century-scale droughts in the Birks, H.J.B., Gordon, A.D., 1985. Numerical Methods in Quaternary northern Great Plains of North America (Yu and Ito, Pollen Analysis. Academic Press, London. 1999). The correlation between North Atlantic drift ice Bond, G., Kromer, B., Beer, J., Muscheler, R., Evans, M., Showers, W., and the Pyramid Lake pollen record suggests that increased Hoffmann, S., Lotti-Bond, R., Hajdas, I., Bonani, G., 2001. Persistent solar irradiance is a possible mechanism to induce extended solar influence on North Atlantic climate during the Holocene. Science 294, 2130–2136. drought in the western Great Basin of North America, Byrne, R., Busby, C., Heizer, R.F., 1979. The Altithermal revisited: Pollen although the link between solar irradiance and climate evidence from the Leonard Rockshelter. Journal of California and Great change is still poorly understood. Basin Anthropology 1, 280–294. Davis, O.K., 1999. Pollen analysis of a late-glacial and Holocene sediment core from Mono Lake, Mono County, California. Quaternary Research 52, 243–249. Conclusion Davis, O.K., Moratto, M.J., 1988. Evidence for a warm dry early Holo- cene in the western Sierra Nevada of California: Pollen and plant The Pyramid Lake pollen record presents a complex macrofossil analysis of Dinkey and Exchequer Meadows. Madron˜o picture of Holocene climate change. The early middle 35 (2), 132–149. Holocene (7600-6300 cal yr B.P.) was the driest period in Davis, O.K., Anderson, R.S., Fall, P.L., O’Rourke, M.K., Thompson, R.S., 1985. Palynological evidence for early Holocene aridity in the southern thelast7600yr,butatleastoneintensewetperiod Sierra Nevada, California. Quaternary Research 24, 322–332. interrupted the general aridity. By 5000 cal yr B.P., maxi- Edlund, E.G., 1996. Late Quaternary Environmental History of Montane mum aridity ended and a period of variable climate fol- Forests of the Sierra Nevada, California. Ph.D. dissertation, University lowed, including clearly identifiable wet and dry phases. of California, Berkeley. Sometime after 3430 but before 2750 cal yr B.P., climate Faegri, K., Iversen, J., 1985. Textbook of Pollen Analysis, fourth ed. Haf- ner Press, New York. became cool and wet, and Lake Tahoe began to spill Grayson, D.K., 1993. The Desert’s Past: A Natural Prehistory of the Great continuously, raising the water-level elevation of Pyramid Basin Smithsonian Institution Press, Washington, DC. Lake. The past 2500 yr have been marked by recurring Harding, S.T., 1935. Changes in lake levels in Great Basin area. Civil persistent droughts. The longest of these droughts occurred Engineering 5, 87–90. between 2500 and 2000 cal yr B.P. Extended droughts Harding, S.T., 1965. Recent Variations in the Water Supply of the Great Basin. Archives Series Report 16, University of California, Berkeley. occurred between 1500 and 1250, 800 and 725, and 600 Hattori, E., 1982. The Archaeology of Falcon Hill, Winnemucca Lake, and 450 cal yr B.P. These droughts correlate with dry Washoe County, Nevada. Nevada State Museum Anthropological periods inferred from ages of submerged stumps and recon- Papers 18. 38 S.A. Mensing et al. / Quaternary Research 62 (2004) 29–38

Hoddel, D.A., Brenner, M., Curtis, J.H., Guilderson, T., 2001. Solar tion of fossil pollen spectra: dissimilarity coefficients and the method of forcing of drought frequency in the Maya lowlands. Science 292, modern analogs. Quaternary Research 23, 87–108. 1367–1370. Shindell, D.T., Scmidt, G.A., Mann, M.E., Rind, D., Waple, A., 2001. Solar Houghton, J.G., Sakamoto, C.M., Gifford, R.O., 1975. Nevada’s Weather forcing of regional climate change during the Maunder Minimum. Sci- and Climate. University of Nevada, Reno. ence 294, 2149–2152. Jankovska´, V., Koma´rek, J., 1982. Das vorkommen einiger chlorokokka- Smith, S., Anderson, R.S., 1992. Late Wisconsin paleoecologic record lalgen in bohmischen spatglazial und postglazial. Folia Geobotanica et from Swamp Lake, Yosemite National Park. Quaternary Research Phytotaxonomica, Praha 17, 165–195. 38, 91–102. Kapp, R.O., Davis, O.K., King, J.E., 2000. Ronald O. Kapp’s Pollen and Solomon, A.M., Silkworth, A.B., 1986. Spatial patterns of atmospheric pol- Spores, second ed. American Association of Stratigraphic Palynolo- len transport in a montane region. Quaternary Research 25, 150–162. gists, College Station, TX. Stine, S., 1990. Late Holocene fluctuations of Mono Lake, eastern Cali- Lanner, R., 1984. Trees of the Great Basin. University of Nevada Press, fornia. Paleogeography, Palaeoclimatology and Palaeoecology 78, Reno, NV. 333–381. LaMarche Jr., V., 1973. Holocene climatic variations inferred from treeline Stine, S., 1994. Extreme and persistent drought in California and Patagonia fluctuations in the White Mountains, California. Quaternary Research 3, during mediaeval time. Nature 369, 546–549. 632–660. Stockmarr, J., 1971. Tablets with spores used in absolute pollen analysis. Lindstro¨m, S., 1990. Submerged tree stumps as indicators of middle Ho- Pollen et Spores 13, 615–621. locene aridity in the Lake Tahoe Basin. Journal of California and Great Tausch, R., Nowak, C., Mensing, S., 2004. Climate change and associated Basin Anthropology 12, 146–157. vegetation dynamics during the Holocene: the paleoecological record. Madsen, D., Rhode, D., Grayson, D., Broughton, J., Livingston, S., Hunt, In: Chambers, Jeanne C., Miller, Jerry R. (Eds.), Great Basin Riparian J., Quade, J., Schmitt, D., Shaver III, M., 2001. Late Quaternary envi- Ecosystems: Ecology, Management and Restoration. Island Press, Cov- ronmental change in the Bonneville Basin Western USA. Palaeogeog- elo, CA, pp. 24–48. raphy, Palaeoclimatology, Palaeoecology 167, 243–271. Thompson, R.S., 1992. Late Quaternary Environments in Ruby Valley, Meko, D., Therrel, M., Baisan, C., Hughes, M., 2001. Sacramento River Nevada. Quaternary Research 37, 1–15. flow reconstructed to A.D. 869 from tree rings. Journal of the American Wigand, P.E., 1987. Diamond Pond, Harney County, Oregon: vegetation Water Resources Association 37, 1029–1039. history and water table in the Eastern Oregon desert. Great Basin Nat- Mensing, S., 2001. Late-Glacial and early Holocene vegetation and cli- uralist 47, 427–458. mate change near Owens Lake, eastern California. Quaternary Re- Wigand, P.E., Mehringer Jr., P.J., 1985. Pollen and seed analyses. In: search 55, 57–65. Thomas, D.H. (Ed.), The Archaeology of Hidden Cave. Anthropolog- Mensing, S.A., Southon, J.R., 1999. A simple method to separate pollen for ical Papers of the American Museum of Natural History, vol. 61. 108– AMS radiocarbon dating and its application to lacustrine and marine 124 (Chap. 9). sediments. Radiocarbon 41, 1–8. Wigand, P.E., Rhode, D., 2002. Great basin vegetation history and aquatic Miller, J., Germanoski, D., Waltman, K., Tausch, R., Chambers, J., 2001. systems: the last 150,000 years. In: Hershler, R., Madsen, D.B., Currey, Influence of late Holocene hillslope processes and landforms on modern D.R. (Eds.), Great Basin Aquatic Systems History. Smithsonian Con- channel dynamics in upland watersheds of central Nevada. Geomor- tributions to the Earth Sciences, vol. 33. Smithsonian Institution Press, phology 38, 373–391. Washington, D.C., pp. 309–368. Moore, P.D., Webb, J.A., 1978. An Illustrated Guide to Pollen Analysis. Wilson, D., 1992. Sawdust Trails in the Truckee Basin: A History of Lum- Wiley, New York. bering Operations Nevada County Historical Society, Nevada City, CA. National Climatic Data Center, 2003. Northern California Climate Sum- Yu, Z., Ito, E., 1999. Possible solar forcing of century-scale drought fre- maries, Station 042467 (Donner Memorial State Park, CA) and Sta- quency in the northern Great Plains. Geology 27, 263–266. tion 048758 (Tahoe, California), available at: http://www.wrcc.dri.edu/ Yuan, F., Linsley, B.K., Lund, S., McGeehin, J.P., 2004. A 1200 year record summary/climsmnca.html. of hydrologic variability in the Sierra Nevada from sediments in Walker Overpeck, J.T. , Webb III, T., Prentice, I.C., 1985. Quantitative interpreta- Lake. Nevada, Geochemistry Geophysics Geosystems 5, 1–13.